ABSTRACT. Granulocyte-macrophage colony-stimulat- ing factor (GM-CSF) and interleukin 3 (IL-3) are pleiotropic hemopoietic growth factors whose genes are ...
Proc. Natl. Acad. Sci. USA Vol. 90, pp. 2466-2470, March 1993 Immunology
The granulocyte-macrophage colony-stimulating factor/ interleukin 3 locus is regulated by an inducible cyclosporin A-sensitive enhancer PETER N. COCKERILL*, M. FRANCES SHANNON, ANDREW G. BERT, GREGORY R. RYAN, AND MATHEW A. VADAS Hanson Centre for Cancer Research, Institute of Medical and Veterinary Science, Frome Road, Adelaide 5000, Australia
Communicated by J. F. A. P. Miller, November 30, 1992
Granulocyte-macrophage colony-stimulatABSTRACT ing factor (GM-CSF) and interleukin 3 (IL-3) are pleiotropic hemopoietic growth factors whose genes are closely linked and induced in T lymphocytes in a cyclosporin A (CsA)-sensitive fashion. Since we found that the human GM-CSF and IL-3 proximal promoters were not sufficient to account for the observed regulation of these genes, we mapped DNase I hypersensitive sites across the GM-CSF/IL-3 locus in the Jurkat human T-celi line to identify additional regulatory elements. We located an inducible DNase I hypersensitive site, 3 kb upstream of the GM-CSF gene, that functioned as a strong CsA-sensitive enhancer of both the GM-CSF and IL-3 promoters. Binding studies employing Jurkat cell nuclear extracts indicated that four sites within the enhancer associate with the inducible transcription factor AP1. Three of these AP1 elements lie within sequences that also associate with factors resembling the CsA-sensitive, T cell-specific transcription factor NFAT. We provide additional evidence suggesting that an APl-like factor represents one of the components of NFAT. We propose that the intergenic enhancer described here is required for the correctly regulated activation of both GM-CSF and IL-3 gene expression in T cells and that it mediates the CsA sensitivity of the GM-CSF/IL-3 locus.
Granulocyte-macrophage colony-stimulating factor (GMCSF) and interleukin 3 (IL-3) are tightly regulated cytokines that direct the proliferation, differentiation, and function of a variety of hemopoietic cells and their precursors (1). Their genes are closely linked; they are just 10 kb apart on human chromosome 5 (2). Both genes are expressed in most T lymphocytes and T-cell lines following activation of the T-cell antigen receptor and other surface molecules (1). The T-cell-receptor-dependent induction of GM-CSF, IL-3, and several other cytokine genes can be blocked by the powerful immunosuppressant cyclosporin A (CsA) (3, 4). The mechanism of action of CsA is well documented only for the interleukin 2 (IL-2) gene, which has a promoter that is strongly induced during T-cell activation (5-9). CsA blocks signals transmitted by the T-cell receptor, which mediate the induction of speciflc transcription factors such as NFAT that associate with the IL-2 promoter. NFAT is a T cell-specific CsA-inhibitable transcription factor induced early in the course of T-cell activation; it is essential for the activation of the IL-2 gene (6-9). The GM-CSF and IL-3 genes, however, appear to be subject to a more complex mode of regulation than IL-2. We observed that the proximal promoters of these genes responded poorly to signals normally delivered via the T-cell receptor and were relatively insensitive to CsA (see below). This suggests that a distal regulatory element(s) may
participate in the activation of the GM-CSF/IL-3 locus and be a target for CsA inhibition. Transcriptional regulatory elements frequently adopt altered chromatin structures, termed DNase I hypersensitive (DH) sites, which can be detected within nuclei on the basis of their enhanced accessibility to DNase I (10). DH sites most likely represent nucleosome-free regions where specific DNA binding proteins associate with DNA and exclude nucleosomes. In this study, we mapped DH sites across the human GM-CSF/IL-3 locus and located an inducible DH sitet in the intergenic region, which functioned as a powerful CsA-inhibitable enhancer.
MATERIALS AND METHODS Plasmid Construction. pHGMO.6 was created by inserting the multiple cloning site AGCTGATATCTCGAGATC-
TGGGCCCGGATCCTGCAGGCCTAAGCTTCGCGAGGTCACC into the HindIII site at the 5' end of the 0.6-kb fragment of the human GM-CSF promoter in the chloramphenicol acetyltransferase (CAT) reporter gene plasmid pCSFp1+ (11). Subsequent derivatives of pHGMO.6 incorporating additional fragments of the human GM-CSF/IL-3 locus contained 2.7-kb Bgl II/HindIIl (pHGM3.3), 2.5-kb BamHI/ HindIII (pHGM3.1), 2.0-kb Bgl II/HindIII (pHGM2.6), or 0.7-kb Bgl II (pHGMB716) fragments of AJ1-16 DNA (2) inserted into the corresponding sites upstream of the promoter in pHGMO.6. pHGMN3 contained three head-to-tail copies of the GM550 element GATCTCTTATTATGACTCTTGCTTTCCTCCTTTCA in the Bgl II site of pHGMO.6. pHIL3 contained a 5.2-kb HindIII/Ban II fragment of A-66 DNA (a gift from S. Clark and Y.-C. Yang) from upstream of the human IL-3 gene inserted into the HindIII/Sal I sites of pBLCAT3 (12). pHIL3B716 had the 716-bp Bgl II fragment inserted into the HindIII site of pHIL3. All DNA fragments were inserted in the same orientation relative to the promoters as they exist in the GM-CSF/IL-3 locus. DH Site Analyses. Nuclei isolations and DNA digestions were performed essentially as described (5). Briefly, nuclei were isolated from Jurkat cells by lysis in 0.1% Nonidet P-40 and resuspended to -0.4 mg/ml(optical density at 260 nm of an aliquot in 1 M NaOH = 10 absorbance units/ml). Aliquots of nuclei were digested for 3 min at 22°C with DNase I (Worthington) at 2-20 units/ml in nuclei isolation buffer containing 1 mM CaCl2. DNA was purified, digested with EcoRI, electrophoresed on a 0.8% agarose gel, blotted onto a Hybond N membrane (Amersham), and hybridized to the 32P-labeled 1.4-kb BamHI fragment of AJ1-16 DNA indicated Abbreviations: GM-CSF, granulocyte-macrophage colony-stimulating factor; IL-2 and IL-3, interleukins 2 and 3; CsA, cyclosporin A; DH, DNase I hypersensitive; CAT, chloramphenicol acetyltransferase; PMA, phorbol 12-myristate 13-acetate. *To whom reprint requests should be addressed. tThe sequence reported in this paper has been submitted in the GenBank data base (accession no. L07488).
The publication costs of this article were defrayed in part by page charge payment. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. §1734 solely to indicate this fact. 2466
Immunology: Cockerill et al.
Proc. Natl. Acad. Sci. USA 90 (1993)
in Fig. 1. One representative sample was selected from each DNase I titration. The extent of DNase I cutting at the DH site was determined by quantitation of 32P emissions with a Molecular Dynamics PhosphorImager. Analyses of mRNA Expression. GM-CSF and IL-3 mRNA accumulation, relative to 3-actin, was determined by RNase protection assay (13). SP6 RNA polymerase transcription templates for the preparation of 32P-labeled RNA probes used in RNase protection assays contained the terminal 207 bp of the human GM-CSF gene (gift from G. Goodall), the 301-bp Sma I/Pst I fragment of the human IL-3 gene, and the 130-bp Pst I/Sma I fragment of the human f-actin gene. mRNA levels were quantitated with a Molecular Dynamics PhosphorImager. The levels of IL-3 and GM-CSF mRNA were normalized as the ratio over the level of 3-actin detected in each assay and displayed as a percentage of the maximum level induced in each series of experiments. Cell Transfections. Jurkat cells (5 x 106) were transfected by electroporation with 10 ,ug of DNA and cultured for 16-24 h before stimulating with phorbol 12-myristate 13-acetate (PMA) at 20 ng/ml plus 2 ,uM A23187, in the presence or absence of 0.1 ,uM CsA. After 6 h the cells were washed to remove the stimulus and cultured for 16-24 h in fresh medium before harvesting cytoplasmic protein for CAT assay (14). Relative CAT activity and standard errors were calculated from at least four experiments by using unstimulated pHGMO.6 set at a value of 1 as a reference. At least two independently derived clones of each construct were tested. The fold induction in response to PMA and A23187 was calculated relative to each unstimulated construct. Gel Electrophoretic Mobility Shift Assays. Mobility shift assays were performed as described (15) using 4% polyacrylamide gels in 25 mM Tris borate/0.5 mM EDTA. Nuclear extracts were prepared as described (16) from Jurkat cells with and without stimulation for 3 h with PMA at 20 ng/ml and 2 AM A23187. In each assay, 4 ,ug of nuclear protein, 5 ,ug of poly(dI-dC), and 0.2 ng of DNA probe were incubated for 25 min at 22°C in 18 Al of 30 mM NaCl, 30 mM KCl, 10% glycerol, 1 mM EDTA, 1 mM dithiothreitol, 20 mM Hepes (pH 7.9), 0.1 mM phenylmethylsulfonyl fluoride, aprotinin at 5 ,g/ml, and leupeptin at 5 ,ug/ml. Assays in Fig. 5b were performed in the presence and absence of 25 ng of specific a
competitor DNA. GM170, GM330, GM400, and GM550 probes and competitors corresponded to the DNA segments indicated in Fig. 2b. The human IL-2 gene NFAT probe and competitor (gift of G. Crabtree) encompassed the sequence GGAGGAAAAACTGTTTCATACAGAAGGCGT. The human stromelysin gene AP1 probe and competitor encompassed the sequence GCAAGGATGAGTCAAGCTGCGGGTGATCC. The nonspecific DNA competitor had the sequence TCGCCAATGAGCTCCCGGGTCGACTGCAGAAGCTTC. RESULTS An Inducible DH Site Lies Upstream of the GM-CSF Gene. Elements required for the correct regulation of the GM-CSF/ IL-3 locus were sought by mapping DH sites (10) in the Jurkat human T-cell line before and after stimulation with PMA and the Ca2+ ionophore A23187. These stimuli mobilize protein kinase C and increase free cytoplasmic Ca2+, respectively, and thus mimic at least some aspects of T-cell receptor activation (1). We mapped DH sites across -40 kb of DNA extending from an EcoRI site 12 kb upstream ofthe IL-3 gene to a Bgl II site 12 kb downstream of the GM-CSF gene. Just one inducible DH site, located 2.8-3.0 kb upstream of the GM-CSF gene, was detected in Jurkat cells (Figs. la and 2a); this site was also inducible in peripheral blood T cells (not shown). In addition, at least seven constitutive DH sites were detected in the vicinity of the IL-3 gene (Fig. 2a; data not shown). Maximal induction ofthe DH site upstream ofthe GM-CSF gene required two signals, provided here by PMA and the Ca2+ ionophore, as did the induction of GM-CSF and IL-3 gene mRNA accumulation (Fig. la). The DH site appeared 1-2 h before the onset of GM-CSF and IL-3 mRNA accumulation (Fig. lb), and its induction was blocked by both the transcription inhibitor dichlorobenzimidazole riboside and the translation inhibitor cycloheximide (Fig. la). Significantly, the immunosuppressant CsA, which inhibited both GM-CSF and IL-3 transcription, also blocked induction of the DH site (Fig. la). The DH site required ongoing induction of nuclear factors for its continued presence, since a significant decrease in intensity occurred 4 h after either addition of CsA or removal of the stimulus (Fig. lc).
b
M
2467
u
c
Q
C4
:
-~
C
+ J-Nr
DHI site
DH
site
00
+
a. r
* s t
RNA 2
GM-CSF 0 0 1L3
1000 0
0
0.3 1.5 8.6 100 0.8 17 1.2
4 6 8 10 12 14 16 hours stimulation + P/I
% max DNaseI 75 100 40 33 site cutting
FIG. 1. An inducible DH site exists upstream of the GM-CSF gene. (a) Induction of the DH site and mRNA expression in Jurkat cells following a 4-h incubation with the indicated stimuli. The upper band is the intact 9.4-kb EcoRI fragment encompassing the GM-CSF gene, while the lower band is a 3.9-kb subfragment generated by DNase I digestion (see Fig. 2a). The numbers below indicate relative accumulation of GM-CSF and IL-3 mRNA in the same cultures. (b) Time course of induction of the DH site (o) and GM-CSF (.) and IL-3 (A) mRNA accumulation after stimulation with PMA and Ca2+ ionophore. (c) The stability of the DH site following induction was tested by either adding CsA after 4 h of an 8-h PMA/Ca2+ ionophore stimulation period or by removing the stimulus at 4 h and incubating for an additional 4 h in medium alone (nil). The first lane provides a measure of the extent of DH site formation after just 4 h of stimulation. 0, unstimulated; P, PMA at 20 ng/ml; I, 2 ,uM Ca2+ ionophore A23187; DRB, 100 ,uM dichlorobenzimidazole riboside; CHX, cycloheximide at 20 ,ug/ml; CsA, 0.1 ,uM. Inhibitors were added 10 min before PMA and A23187.
2468
Immunology: Cockerill et al.
a
b
Proc. Natl. Acad. Sci. USA 90 (1993)
constitutive DH sites
'i
i
f
f
I
I
IL3 RR _I--W
Hl
B
B
B
R
I,
I
R
-
GM-CSF G
I I 111
I
I
I
II
X
B
XB
H
I
I
B
B
2.2 kb
2.9 kb XhoI/StuI
5.2 kb HindIII/BanII pHIL3
inducible DH site G iGG
probe
I
.I_ -
R
I FW-
I.-
H IpHGM3.3
BamHI iI
716 bp enhancerl-I
IpHGM3.1 IpHGM2.6 I-i pHGMO.6
AGATCTC-LAGGTCCCCUAGAGATAAAAGAACCCTCrATCAAGGGTTGGACAGGTCAAGAGGATCAGATCAGCCCCAAAGCCTGTGCGGGGGTGCCATGGTACCGAAGGACTTTGAGGCCC 120 BglII GM170
ATTTCCTTCCCCTTGTGATAATGTCTCTCGTGATAAGGATCCTGGAGTGACTCAAGCCCCTGTTTCCTACAGAGCCTGTGGGGTGGACACGCATAGGAAACTCCTTCCAGAGGGTTTTCT BamHI
240
CCTGTCTCTGTAGGAAGGGGGGCCCCAGAGGGGTCATACCACCCATCGTGGTCCCCGCCTCTGCCCTGCCACAGCCCCATCGGAGCCCCTGAGTCGCATGGCTGGCTATCGGTTGACAC
360
AP1
ApaI
GM330
APi
GM400
TGTTTAAAATACAGTGCAATCTAGGAATGCCTGCCTGTGTCATGGCCTCAGCCCTGATGTCATCTTTCQATGAGAAAGATGTGTGGCTGCCCACAAACCTCATCACCAAACCCQTGGATT API GM550
480
TCTGTCCTGTTTGTGCTGACTTGGGCACCTGCAGGCCACCACTGCCGAAATGGATTTCTTATTATGACTCTTGCTTTCCTCCTTTCCTTCCCTCCTAGGCCCTTGGCCACTGGCCTGTTA
600
CTTGCCCATCTGTTATGTCCCAGCCCTTAAACACAGCTTTGCTGGCCCTCAGCTCTTCTTCCTGAATAACTTCCGTGGCCTCAGCTACCACACAGACATGTACCGCTGCCAGATCT BglII
716
PstI
P
FIG. 2. Location and sequence ofthe inducible DH site. (a) Restriction enzyme map of the human GM-CSF/IL-3 locus displaying the location of DH sites (vertical arrows). The nmap was derived from the phage clones AJ1-16 (2) and A-66 (obtained from Y.-C. Yang and S. Clark) and the plasmid pCH5.2 (11). The solid bar above the map represents the DNA probe used to map the inducible DH site. The bars below indicate DNA fragments tested in subsequent transcriptional analyses. H, HindIII; B, BamHI; R, EcoRI; X, Xho I; G, Bgl II. (b) DNA sequence of the 716-bp Bgl II fragment spanning the DH site. APl-like elements are underlined, whereas DNA probes and competitors used in subsequent electrophoretic mobility shift assays are indicated above the sequence. The DH site lies between the Apa I and Pst I sites. The DNA sequence was determined using an Applied Biosystems automated sequencer, by sequencing each strand of overlapping DNA clones derived from AJ1-16.
GM-CSF and IL-3 Are Regulated by an Inducible CsASensitive Enhancer. The inducible DH site was tested for transcriptional regulatory activity in transient transfection assays in Jurkat cells. Segments of up to 3.3 kb of DNA, extending upstream from the GM-CSF gene transcription start site, were linked to the CAT reporter gene (pHGMO.6pHGM3.3; Fig. 2a). DNA segments that include the DH site (pHGM3.3 and pHGM3.1) had little constitutive activity, but mediated an 80- to 95-fold induction of CAT in response to PMA plus Ca2+ ionophore (Fig. 3a). In contrast, very weak induction of CAT was supported by GM-CSF promoter segments lacking the DH site (pHGM2.6 and pHGM0.6; Fig. 3a). Furthermore, a 716-bp Bgl II fragment of DNA encompassing the DH site functioned as a powerful inducible enhancer when linked directly to the 0.6-kb GM-CSF proximal promoter fragment in both its normal (pHGMB716; Fig. 3b) and reverse orientation (not shown). No additional enhancers were detected upon testing a 2.2-kb BamHI fragment immediately upstream of the inducible DH site or a 2.9-kb Xho I/Stu I fragment encompassing three constitutive DH sites just downstream of the IL-3 gene (Fig. 2a; data not shown). The enhancer was also required for efficient induction of the IL-3 promoter in transient transfection assays. A CAT gene plasmid containing 5.2 kb of the IL-3 promoter and upstream DNA, including four upstream DH sites, had a moderate basal activity, but was only weakly induced by PMA and the Ca2+ ionophore (pHIL3; Figs. 2a and 3b). However, a reduced basal activity and a 10-fold induction of CAT was obtained when the 716-bp enhancer was coupled to the IL-3 promoter, 5.2 kb upstream of the CAT gene (pHIL3B716; Fig. 3b). Similar results were obtained when the enhancer was placed either upstream of the promoter or downstream of the CAT gene in an IL-3 promoter/CAT gene construct containing just the 315-bp Stu I/Ban II fragment from immediately upstream of the IL-3 transcription start site (data not shown). The enhancer appeared to be the major target for CsA in the GM-CSF/IL-3 locus, since CsA suppressed the enhancer-driven activation of both promoters (pHGMB716 and pHIL3B716; Fig. 3b). In contrast, CsA had little inhibitory effect on the GM-CSF promoter alone (pHGMO.6; Fig. 3b)
and unexpectedly led to a slightly increased induction ofboth the 5.2-kb (pHIL3; Fig. 3b) and 315-bp (data not shown) IL-3 promoter constructs used in this study. The Enhancer Binds AP1 and NFAT. The 716-bp enhancer fragment was sequenced and found to encompass four consensus binding sequences for the inducible CsA-resistant transcription factor AP1 (17, 18) (Fig. 2b). Two of these sites (located at positions 170 and 330) match perfectly the TGAGTCA AP1 consensus, while the other two (located at positions 400 and 550) have a single base mismatch. The AP1 element at position 550 is, however, identical to a functional AP1 element in the IL-2 promoter (19, 20) (Fig. 4). Furthermore, this AP1 element is contained within a broader region (segment GM550; Fig. 2b) homologous to an NFAT site located in a more distal region of the IL-2 promoter (Fig. 4). NFAT could, therefore, also participate in the CsA-sensitive activation of the GM-CSF and IL-3 genes in Jurkat cells. Indeed, three linked copies of the GM550 segment functioned as a strong inducible enhancer inhibited by CsA, when placed upstream of the GM-CSF promoter (pHGMN3; Fig. 3b). To identify the transcription factors that mediate enhancer function, we assayed fragments of the enhancer in electrophoretic mobility shift assays with Jurkat cell nuclear extracts prepared from unstimulated cells and cells stimulated in the presence and absence of CsA (Fig. Sa). Complexes induced by PMA and Ca2+ ionophore were formed with four DNA segments encompassing AP1 sites (GM170, GM330, GM400, and GM550 as in Fig. 2b). Each probe formed a CsA-resistant complex that comigrated with a stromelysin gene (21) AP1 complex. Furthermore, all four AP1 sites were protected by recombinant c-jun protein (18), a component of AP1, in DNase I footprinting assays (data not shown). Three of the probes (GM170, GM330, and GM550) also formed inducible CsA-sensitive complexes that comigrated with an inducible complex formed with the distal NFAT site from the human IL-2 gene (6, 8, 22). The IL-2 NFAT probe formed two inducible CsA-inhibitable complexes similar to those detected with a mouse IL-2 gene NFAT site (22). The lower of these complexes comigrated with AP1 and may be a previously described component of NFAT that lacks AP1 (22). All
_
Immunology: Cockerill et al.
Proc. Natl. Acad. Sci. USA 90 (1993)
a.
40
GM170
GJM330)
U
30
NF-AT'
API
GM550
I
1
-
GM400
2469
OaX C
oin
5,
cr,o
u
C-
a"X
,Z_L
u:
uz
C
C
CL
L
c-L:
c
XL-
-L
l
H> u 2-
NEAr
20
0
W
-*
""I. " Wm"
10
FOLD
*
&
'AX4
HGM3.3
JfA
HGM3.1
_
_
HGM2.6 HGMO.6 00
m
INDUCTION: oc 50
b | *NIL
40
-
>
/
/f -0
30/ 20
b
GM170
GM33()
APi-
GM55()
NFAl'I
--
1. LL .sD
NFATl-
GM40()
zI zi-
-
